Why do we spend all this time avoiding the truth? We have antitrust laws to dismantle monopolies. This is because monopolies can be dismantled by two things:
1. The coercive power of the state.
2. A sudden realization among those who run and benefit from the monopoly that selfless love for their fellow human beings is enough reason to stop predatory rent-seeking.
In other words, only love or power. We spend a lot of time not talking about the two main forces that guide human life. Not reason. Not reformed open source software development. Love and power.
I often suspect this reluctance to involve the coercive power of the state comes from a libertarian streak in the tech world. But I am a weird anarchist kind of guy so maybe I am mistaken on this.
As the article touches on, we spend the time because people want to address companies like Facebook, which are not obviously monopolies in the normal legal sense. They don't overcharge users, they have strong competitors, etc. But they're still a huge tech company with outsized influence on both society and software development industry.
I’d argue they are a monopoly in the traditonal sense too. They control the whole stack like standard oil. They probably do overcharge their actual users/customers (the advertisers, not you, the product they sell access to). The more you read about places like standard oil the more they reflect these modern companies.
> This is because monopolies can be dismantled by two things:
> 1. The coercive power of the state.
> 2. A sudden realization among those who run and benefit from the monopoly that selfless love for their fellow human beings is enough reason to stop predatory rent-seeking.
You're missing the third option: Someone else finds a way to break their lock on the market.
In many cases this happens indirectly. If you had a monopoly on local media before the internet, you don't have one after the internet.
But that doesn't mean you can't do it with a purpose.
In many cases this requires an act of altruism. To break the monopoly, somebody has to spend non-trivial resources to reproduce the ability to do something that currently only the monopolist can, without hope of turning a profit from it. People can do this if they choose to.
And the work required to do it can be distributed throughout a community.
It’s humorous to me that we discuss “monopolies” over what are ultimately very discretionary sectors. Compare to the trusts of the 20th century, which were monopolizing oil or steel for example. These are (or were) the commodities required to power commerce via shipping and transportation. A single entity having pricing power over such essential commodities threatens the robustness of the world economy.
Today we complain about the monopolistic power of, e.g. Google, who does not actually have a monopoly on search (there are many other providers), and the ability to search the entire internet isn’t really that necessary, or Apple’s mobile marketplace, where they again do not have a monopoly (people choose to develop for Apple devices, and part of choosing to sell something somewhere is agreeing to the market’s terms - as has been the case for all of time), and this is mostly to power manipulative microtransactions for kids on games. These are entirely different ballparks.
The thing you're missing is the "you're not the customer, you're the product."
The customer is the app developer or the advertiser -- the startup Apple is extracting 30% from, or the small business which just wants to get its name out but Google systematically excludes small websites from organic search results and requires you to pay a huge chunk of your revenue to be seen.
If your prospective customers have iPhones, you need Apple's permission or you're out of business. If your prospective customers use Google search, you pay them for advertising or you're invisible. You don't get to choose which phone or search engine they use, and they rarely if ever use more than one -- unlike retail stores where it's common for the same customer to shop at dozens of different stores.
If Walmart won't carry your products in their stores, you can sell them through Target and Amazon and a zillion other retailers. The same customers can get it from another store. If Apple won't carry your app, those customers have no other viable means to get it from you.
This isn't just "manipulative microtransactions for kids on games" -- it's the reason you can't use cryptocurrency or any competing payment network for in-app purchases. Are payments not an important industry for which we care about competition?
Apple basically murdered Tumblr by giving them an impossible choice between banning adult content (causing them to lose too much of their user base) or being banned from iPhones (causing them to lose too much of their user base). Regardless of how you feel about adult content, they have the power to do this over anything. Should they be able to destroy your social network or force you to censor if they don't like your politics? How does this not matter?
They do apply to corporations. Allowing commercial use is part of the definition of both "open source" and "free software" so a non-commercial license is incompatible with the GPL
Without Stallman releasing an official Non-Commercial GPL (NCGPL) and allowing GPL'd software to be upgraded to it like it can be upgraded to AGPL, there's no way to merge non-commercial licensed code into GPL code without relicensing even if the non-commercial code is otherwise using a standard FOSS license.
This article sounds like the author started with a conclusion and worked their way back to a premise. Monopolistic powers are a business and economics problem, it has nothing to do with the building blocks used to build their products.
>For example, if Uber’s driver rating could be exported to other similar ride-hailing services, it would have the effect of reducing the lock-in effect of Uber platform.
People are not using Uber because they don't have a rating on newcomer apps. They're using it because Uber has a large number riders and drivers in many locations, and people are aware of it.
> Mandate the use of open standards and open source competitor: All solutions must have open source alternatives before being implemented. For example, the Bavarian government switched the government operating system to Linux in 2018 to avoid being subject to Microsoft’s monopoly on software. While the project was not entirely successful
They discovered that there's a difference between "software" and "a software implementation". A myopic view of software as just code that you can throw over a fence is missing the entire reason that Windows was and is dominant on desktop. Deploying, integrating, maintaining, and operating software across an organization is labor intensive. Microsoft doesn't "compete with Linux", they compete with Red Hat or Canonical.
> People are not using Uber because they don't have a rating on newcomer apps. They're using it because Uber has a large number riders and drivers in many locations, and people are aware of it.
I think the general point is that you want to move the network effect outside the control of a single entity. So for example, like email rather than like Facebook. You relegate the service into something narrower, like payment processing, which then becomes fungible. You find the driver you want in a decentralized system and then pay them using Uber or Paypal or CashApp or Mastercard or physical cash or whatever you like.
> Deploying, integrating, maintaining, and operating software across an organization is labor intensive. Microsoft doesn't "compete with Linux", they compete with Red Hat or Canonical.
That depends on your scale, but there are different ways to go about it.
If you're small, just compiling something from github is a reasonable thing to do. You're not deploying it to a thousand workstations or trying to centrally manage anything. What do you need with Active Directory?
If you're large, you have a well-staffed IT department and possibly even your own software developers who can handle deployment issues in-house. Google has its own in-house Linux distribution in addition to ChromeOS.
The target market for Windows and Red Hat and Canonical is medium-sized entities with enough money to pay someone else to solve their problems but not enough to bring that expertise in house.
But many corporations and governments are large enough that they'd be better off doing what Google does, and open source communities can operate at the same scale. Medium sized companies that actually have a dedicated IT department could very well be better off to make a minor contribution to a community effort, and use it themselves, than lock themselves into one of these enterprise vendors. That would require more organization than is currently present, but not necessarily more than is humanly possible.
> You relegate the service into something narrower, like payment processing, which then becomes fungible. You find the driver you want in a decentralized system
But how? Who is going to build, maintain, and provide equitable and effective governance, support, and regulatory compliance for that system? And why are others going to use it?
> If you're small, just compiling something from github is a reasonable thing to do.
Most small orgs don't have IT expertise. They buy off the shelf hardware and software. The computers they buy come with an OS on them, and they use that one. What would compel them to install a different one?
> If you're large, you have a well-staffed IT department and possibly even your own software developers who can handle deployment issues in-house. Google has its own in-house Linux distribution in addition to ChromeOS.
Google is a software company, of course they develop software. Most organizations are not software companies. There's really no reason for other large companies like PepsiCo to be making their own Linux distros for their office staff, any more than it makes sense for Google to start bottling their own beverages for their cafeteria. Yes, many are large enough that they could -- but they'll never do as good as a software company does, because they aren't one.
The question isn't whether they could do it, the question is: are there enough benefits to outweigh the costs and risks? Are you really better off training and supporting your office staff to send out invoices using SumatraPDF on Ubuntu, just to save money on Adobe/Windows licensing? For companies that have nothing to do with software, this is a recipe to lose more money in operational problems than they save in licensing.
> But how? Who is going to build, maintain, and provide equitable and effective governance, support, and regulatory compliance for that system? And why are others going to use it?
Who does it for Debian or Blender or DNS or SMTP?
> Most small orgs don't have IT expertise. They buy off the shelf hardware and software. The computers they buy come with an OS on them, and they use that one. What would compel them to install a different one?
The computer they bought came with Windows 7, but Windows 7 isn't supported anymore. Windows 11 says it doesn't support that hardware. Meanwhile it's a quad core with enough RAM and an SSD and even if it's a few years old it's still fine. All they need it for is to use Google Docs and Chrome/Firefox. Why spend hundreds of dollars on a new machine when you can just have your niece install Debian?
> There's really no reason for other large companies like PepsiCo to be making their own Linux distros for their office staff, any more than it makes sense for Google to start bottling their own beverages for their cafeteria.
There is though.
Bottling beverages is a low margin commodity market. Software isn't. If you can bottle your beverages just as well as Pepsi, you might save a fraction of a percent on something that costs you a fraction of a percent of your revenue. If you can cut Microsoft out of a large enterprise, you can eliminate tens of millions of dollars in licensing costs. That pays for a non-zero number of software developers.
Meanwhile free software exists, so you're not starting from scratch with an empty plot of land and have to figure out how to mold plastic into beverage containers. You start with a Linux distribution that other large enterprises are already using and make some minor modifications to tailor it to your needs.
> Yes, many are large enough that they could -- but they'll never do as good as a software company does, because they aren't one.
In many cases they are. If you're a bank, you're "not a software company", but how are your customers interacting with you in the main? How about a health insurance company? A telecommunications company? An online retailer, like Amazon (which knows this quite well)?
You don't have to be "a software company" for software to be a major part of your business.
> Are you really better off training and supporting your office staff to send out invoices using SumatraPDF on Ubuntu, just to save money on Adobe/Windows licensing? For companies that have nothing to do with software, this is a recipe to lose more money in operational problems than they save in licensing.
It's not just about licensing. Okay, so you save a couple hundred bucks per seat on licensing. That's several million dollars, you could justify that, but maybe it's not much better than breakeven.
But then you're not trapped in a Skinner box, where every day you have to keep clicking "no" to a box that, if you ever click yes, you end up violating a privacy law and putting your company in regulatory peril. Or locking yourself into a vendor who can then raise prices on you by more than your original estimate of what they would cost, but by then it's too late and the cost of extricating yourself is now more than the cost of paying the danegeld.
There is value in opting yourself out of the enshittification of major platforms.
> This article sounds like the author started with a conclusion and worked their way back to a premise.
That’s how all articles or persuasion techniques work! You have a message you want to convey in mind, and then you decide how to convince the reader and write accordingly!
If you rephrase this as fighting consolidation through open source software, I think it is rather clear that you cannot. Consolidation will always have benefits at the small. That there is a straw that breaks that camel's back is not an argument against carrying more straw.
I am personally of the mind that we need to somehow shift the care not from personal future but to collective future. I don't have any thoughts on how to actually do that, though. It is very hard to convince people that are legitimately working hard for their personal well being that they should also give thoughts to those that come next.
> It is very hard to convince people that are legitimately working hard for their personal well being that they should also give thoughts to those that come next.
Sorry - I wasn't clear. I mean what have you tried that leads you to know this:
> It is very hard to convince people that are legitimately working hard for their personal well being that they should also give thoughts to those that come next
I'm mainly reflecting on friends and family that are heavily anti tax. Probably most easily seen by places that have a hard time justifying the funding of a school district.
I knew a guy who was not exactly anti-tax, but he believed that poor people should be taxed more heavily than the rich ones, the moral justification being that that would motivate the poor to try harder to be rich; plus, of course, there are more poor people, and taxing them would likely yield more money than taxing the rich, not to mention that much of the taxes go to the social programs to support the poor, so it's only fair if they paid for those themselves.
I can’t speak for OP but the fact that business ethics classes tend to be viewed not as cautionary tales but as how-tos indicates to me that the whole system is rotten at the core.
One of the kids I know is in a business ethics class right now and thinks it's bullshit ("describe a time when you were a decent human being") so I see things haven't really changed in 25 years, and the old jokes still work.
I’m told that those 80’s and 90’s Greed is Good/Wallstreet movies were spawned by a book called Liar’s Poker. Like the guy who accidentally invented Waterfall, it was meant as a cautionary tale and people thought it sounded good, let’s try it.
> It is very hard to convince people that are legitimately working hard for their personal well being that they should also give thoughts to those that come next.
I think the way to do this is to put their gains in jeopardy. We need to sit down and find an example where some money-making endeavor made life worse for most people, and scrub those dollars from the record. Perhaps we'll do this yearly.
Make dollars gained doing obvious good more reliable than dollars gained doing obvious harm.
The issue is who decides the obvious good and obvious harm? As soon as you’ve made such a powerful authority it becomes a target for regulatory capture since the incentives if you pull this off are sky high.
No need for an authority, it can be a DIY/opt-in sort of thing. If somebody came up to you and presented you with some money that had metadata re: how it was earned, you'd have to decide whether to accept it, so it would be up to you to decide if that endeavor was one you wanted to align yourself with.
If you make that call in a way that disagrees with your wider community, you may end up in a situation where you can't find anyone else who will accept that money, so I expect some degree of consensus will emerge.
Making that judgement call on every transaction would be tedious, so I imagine organizations would form which are trusted to make that call, and then you'd sort of "subscribe" your wallet to their decisions so that you don't have to think about it all the time.
As for regulatory capture, if such an organization stopped aligning with your interests, you could just unsubscribe from their judgements and either do it yourself or form a competing one.
See, I think this is a great reason to colonize Mars or the Moon. Because I'd like to see the result of this social experiment without watching the world crash and burn at the whim of popularity economics.
We have the sky high regulatory capture as it is, just that what has been captured is opaque. There are loads of things that corporations should be getting fined for but are not.
> We need to sit down and find an example where some money-making endeavor made life worse for most people, and scrub those dollars from the record. Perhaps we'll do this yearly.
And then what? Just sitting down and finding out that some thing some people did made some other people's life worse won't put the gains of the first group of people in jeopardy — unless you do something more tangible. What do propose that should be?
Once we know which money is working against our interests, we stop accepting that money. If people show up at the grocery store with it, they leave empty handed. Eventually, people start assessing risks differently, and stop investing in endeavors that are likely to leave them in that situation.
Who are those "we"? How do you check that nobody of those "we" defects? Strikebreaking is a thing.
> If people show up at the grocery store with it
What if those grocery stores are actually the chain that's owned by one of those "bad" companies? Those grocery stores won't implement your policy, and now what? Plus I'm fairly certain you'll run afoul of some anti-discrimination law or another.
Really, if you're engaging in wishful thinking you may as well just wish for everyone to be nice and be done with it.
> Really, if you're engaging in wishful thinking you may as well just wish for everyone to be nice and be done with it.
I disagree. "Everyone be nice" is not something that can be engineered. As hackers, we can live it, and we should, but we can't hack towards it.
> Strikebreaking is a thing.
The kind of information hygiene that makes strike breaking a more difficult endeavor is the kind of thing we can hack towards.
So many of our problems today come from how we've organized our information infrastructure (there's a lot of centralized authority in it). If we find ways to do it differently, if we practice ways to do it differently, then we may be able to achieve different outcomes.
We need a PBS for websites and a PBS for software.
But some of the rules that made PBS PBS have been rolled back so I don’t know that this will happen any time soon. In its place we have various foundations but those aren’t enough.
We're all stuck in the quagmire of Oracle databases, licenses, and lawsuits, because no open source alternatives exist. What a shame that open source hasn't deterred any monopolies and can't possibly do so in the future.
There are no currently coherent checks on this because of regulatory capture. There exist no socialist governing bodies that have power at the same scale as capitalist ones. (Note: please no “might make right” rejoinders)
The us government has set precedent over the last ~40 years which has intentionally, and structurally enshrined the capital class as legally distinct and more protected in practice
I have Google Fonts blocked, so I see the headings on the page displayed using the generic cursive font which ends up being… Comic Sans. Which certainly gives this article a unique kind of gravitas.
A pretty strange choice anyway, given that Maven Pro doesn’t even seem to be a cursive font at all.
I feel like you could make a strong argument that open source actually aids tech monopolies. Consider some business that needs <foo> to work. They can use OSS, buy, or build. If they buy, they share their profits with some other company, and they have to do that for the duration of their use of <foo>. If they build <foo> as proprietary software then they are stuck with keeping expertise in that software. If <foo> is instead OSS, then they can either directly use that component, or make some initial investment in that component, and then use it with minimal further overheads.
If you are building, then it better be your core competence. If you buy, then as you become larger and more successful, buying will be more and more expensive. Its hard to make your suppliers exclusive, and competition will be somewhat maintained.
The Linux Kernel is a great example. There used to be a ton of kernels to choose from maintained and developed by small businesses, but over time Linux was chosen as a good free option. Companies would then put relatively minor investments into customizing or maintaining Linux, avoiding needing a proprietary kernel. After 20 or so years of this, there is almost no money to be made as a small operating systems vendor.
If they are the primary maintainer of <foo> then that's no different than building it in house and it becomes "open source" in the way that Android is, where they throw the code over the wall once a in a while but don't develop it in public. This only helps them in the sense that it prevents a competing ecosystem, by making the one they control free-as-in-beer. But that is something they do, not something you do. Whether or not you produce some other free software yourself isn't going to change it.
If they're using some common community-developed free software, that's good, because they contribute to it without controlling it, and then anyone else can use it too and it's one less component of the stack that someone else has to build or buy in order to compete with them.
I'm not sure how that drives monopolies though. All you've explained is how it lowers the barriers to market, but it does so for all companies, competitors included.
It lowers the barrier to market for the end consumer product, but all of the stuff in between is gone. No more compiler vendors, or xml libraries, etc.
Are you implying that's a bad thing? Your argument sounds an awful lot like, "but all of those middlemen are gone", which is generally considered to be a good thing that results in a more efficient market overall.
Is there some benefit to having compiler vendors or xml libraries guarded by paywalls? This is non-obvious to me. The days before the FSF and gcc were considered "dark ages" to programmers, being locked in to a small set of proprietary of compiler vendors and toolchains, fragmentation everywhere with poor portability and poor hardware compatibility. Yet you paint this as somehow a good thing and I'm struggling to see why.
I'm also still not seeing how the openness helps establish tech monopolies, which was the main point I took issue with.
I agree, to an extent: I think OSS commoditized the substrate of modern technology: OS, libraries, languages, etc, which made the higher-level stuff more valuable (Christensen's law). With commodity platforms, most of the value was left in the end product (i.e. the user-facing bits) and at that point market dynamics favored natural monopolies --the more customers a service had, the better it got, the more customers it got. Which is another reason user-facing OSS apps/services have mostly failed miserably.
It's deeply ironic that in the end OSS, that lovely near-communist idea helped create what is probably the most successful capitalistic business models of all time... but, the fact that these monopolies are still there and unregulated and unharmed, that's probably a US-specific political failure and cannot be fixed with technological hacks and definitely not by one of their core enablers (OSS).
I feel modern "open source" is kind of a parody of the original intentions and it completely misses the forest for the trees.
Wikipedia says the goals of the GPL are "the four freedoms to run, study, share, and modify the software."
It's a pyrrhic victory in today's world where the building blocks are open source, but the full stack is not, and the software is running on corporate servers where users have no control over it.
Companies have simply commoditized and devalued the software and started hoarding the far more valuable user data instead.
FOSS developers should probably slap a license on it that says it's free unless you are a company that has a revenue greater than $X million dollars. This is still fair to everyone since for the company the price of using the software would be peanuts. And now at least these companies are paying for the development and maintenance.
What stops Amazon from establishing a separate company called Totally Not Amazon, which employs 1 person, and hosts your software for the rest of Amazon, and has revenue much less than 1M$ so they can use it for free?
If they need more capacity than 1M would plausibly buy them from themselves, just establish however many more identical companies are necessary, none of which exceed 1M$.
Probably language that suggests if amazon would like to do that they would need to split the entire company into a million pieces to each be under the revenue limit.
In practice though amazon engineers would reverse engineer any tool they cant legally use and need.
Probably the same thing that stops Amazon from establishing a Totally Not Amazon that has zero revenue, so that Amazon can use all sorts of "free for non-commercial use" software.
Ironic that people complaining about free riders are trying to be free riders on the goodwill of "Open Source" by repurposing it to mean something else.
I'll happily pay if my company reaches millions in revenue. Which is only fair since my software is built on top of all kinds of open source software. I don't see the problem of paying only if the software brought me success.
Agpl has no clause for large companies, so it's different.
Also, may I ask what is your fundamental problem with such a license besides perhaps the name and purity? This license will only affect the insanely rich, and in a way that they will most likely not even feel.
License profileration, which makes it harder for people to know what is acceptable or not. Also, GPL is well-known, well-tested, and already excludes most companies. If they want to use our stuff, they better pay it back and become GPL too.
That's the practical concern. The idealistic concern is that free software is all about handing power to the users. Much like how a pen can't restrict how you use it, developers shouldn't be able to restrict how users can use it. Note that the only restriction GPL has is that you cannot further restrict it (tolerance paradox). It's an ethical position to give power to the users, regardless of who they are, because software owners should not have the power to control the users. From here, a software developer decreeing "you may not use my stuff if you make more than X amount per year" is abusing their power over users, which is unethical. Imagine if a pen company told you you cannot use their pens if you met X Y or Z conditions. It's just not right.
Free Software is supposed to create a gift economy. People create and release their work in good faith, others copy and improve, and contribute back to the community. Gift economies work if most continue to gift their work to society in good faith.
If some actors decide to largely take and but rarely give out of self-interest, or worse use their gifts to cement their economic position, then the gift economy turns into an exploitation economy. The way to fix such an economy is to intentionally exclude those who are exploitative.
Free Software failed to create a gift economy. I am all for abandoning it in favor of the Guarded Free Software economy as you propose.
In a regular economy, every exchange is a two-way transaction. I give you something, you give me something (usually of perceived equal value).
In a gift economy. I gift you something. You gift someone something. That someone could be me, but if I don't need/want anything you need, then you just gift to someone else. If there are gifts going around all the time, all in good faith, then most people will receive what they need.
You see gift economies in families or friends all the time. People help others according to their abilities. My rich friend takes me out to an expensive dinner, poor me gets us lunch from that famous food truck. I nurse my uncle when he breaks his leg. He lets my brother's son to live with him for free during college. There is no accounting of how much anyone gave. Just give what you can.
I recommend reading Cory Doctorow's novel Walkaway which describes gift economies beautifully.
That's not free though. That's something very similar to "free for non-commercial use" (with an extension allowing free commercial use up to a certain scale).
It doesn't solve the problem of someone being able to operate the software on a server, where the users have no control of it and their data is held hostage, too.
I think you're confusing open source and free software. Open source was never about the full stack. It was exactly what you found in Wikipedia.
The issue today is that people want more than open source ever promised. The source is available, you can make your own changes, you can distribute them or even start a business but that's not enough for most people/companies. They want a ecosystem, a non-profit organizing the effort, ability to buy support, etc. None of that were in the initial goals.
Only if you’re not in the SaaS space. If you are, then open-sourcing your tech is a risky strategy because then if you have any success, you start to compete with bigger players who take what you’ve invested & then it’s a game of $ which is harder to compete with.
It can still be open source. The AGPL was designed for protections in that scenario. Unless you’re suggesting the AGPL is insufficient. There’s another one that’s really similar, too.
I don’t know many software developers that are OK using AGPL software. The license will infect all products that someone may try to go build even if it doesn’t compete.
MongoDB and Elastic famously didn't find it to be sufficient, swapping to a new and non-OSI-approved license to protect against big cloud providers offering managed versions of their software.
When has open sourcing your business's core software not been a risky strategy?
If you are a software as a product company and you open source your software, someone else may put it into a box and sell it beside you.
If you are a software as a business process company and you open source your software, some else may build into their business and compete against you in whatever your marketable output is.
If you are a software as a service company, and you open source your software, someone may build a service that is remarkably similar to yours.
Open source is great for things your company depends on --- maybe you get community improvements, and you can build and change it how you like; but if the core of your business is open source, that's a choice that makes it very easy for competition to form. OTOH, if you want to build something that other companies depend on, and there's open source alternatives, it's going to be hard going; I wouldn't use a closed source web server or database server or search engine, given there are good open source options; although maybe if those options don't work out, I'd consider a closed source vendor.
> The source is available, you can make your own changes, you can distribute them or even start a business but that's not enough for most people/companies.
You really can't in the original sense, though. A person using Facebook, Spotify, or YouTube can't make any modifications to those products whatsoever despite a lot of the code being "open source". It was never about taking somebody else's code and making a business.
More disappointing, users often don't even have direct access to their own data on those services.
Ok, but Free Software has also failed at the stated goals—we still have no control over the software we use even if it's copyleft software backing it. This is a useless comment.
At this stage the weaponization/enshittification of FOSS by abstracting its original goals, and making it about the equivalent of leverage for larger firms to use for cheap labor/R&D will only likely change if people only use reciprocal licenses with clauses for SaaS firms to straighten up (e.g. AGPL v3) AND legally enforce them in the courts.
I think the key is data interoperability and open protocols.
Ideally peer-based but I know federation is much more popular and that can work also.
But it's not just the fact that they leverage open software to build their closed systems, it's the large networks and marketplaces they have. The only way to compete with that is through decentralized systems. Which (despite the connotation of the name), are the best approaches for creating large scale, open, holistic networks.
Moving to decentralized protocols also means it is harder to dominate with one piece of software. Ideally the protocols are in some kind of protocol registry so they can evolve freely.
But on some level I believe that some entity like a government needs to mandate this. Although I know there are many people who are philosophically against governments or practically against them due to their track record of being generally awful. But at some level you may have to force people to cooperate in some basic technical way in order to avoid the technopolies.
I think decentralized networks essentially provide another opportunity for businesses. Look at email, or web browsers and how centralized they are now.
Make an app for a decentralized protocol, add some polish and maybe give something away with it which is cheap at scale (storage, bandwidth) as the loss leader and establish dominance.
To me just like driving there should be a policy that mandates people respect & act in a way that both you & everyone can move freely and comfortably, lightly, this would be an actual order. We need something like alien philosophy/religion and let us be ruled by it, or, like, everybody just start acting japanese.
We just need to mandate the basic / baseline technology. External File System that is supported everywhere ( ExFAT ? ), PDF, Word, Excel / Spreadsheet, Archive Files, Photo/ Graphics, Audio and Video.
Unfortunately I dont it will ever happen. One being goverement doesn't think like that. Two even if they do they dont know which one to pick. And three since they cant pick anyone they will decide to make their own.......
Right now I really wish we could create something on top of the soon patent free AAC-LC and H.264 Video Codec.
Uh, lol, no. Software development has nothing to do with business plans. Don't like monopolies? That's economic policy, not software. You can't code your way out of politics. But mandating legislation of open source is a great way to kill it if that's your goal.
111 comments
[ 0.19 ms ] story [ 184 ms ] threadWhy do we spend all this time avoiding the truth? We have antitrust laws to dismantle monopolies. This is because monopolies can be dismantled by two things:
1. The coercive power of the state.
2. A sudden realization among those who run and benefit from the monopoly that selfless love for their fellow human beings is enough reason to stop predatory rent-seeking.
In other words, only love or power. We spend a lot of time not talking about the two main forces that guide human life. Not reason. Not reformed open source software development. Love and power.
I often suspect this reluctance to involve the coercive power of the state comes from a libertarian streak in the tech world. But I am a weird anarchist kind of guy so maybe I am mistaken on this.
> 1. The coercive power of the state.
> 2. A sudden realization among those who run and benefit from the monopoly that selfless love for their fellow human beings is enough reason to stop predatory rent-seeking.
You're missing the third option: Someone else finds a way to break their lock on the market.
In many cases this happens indirectly. If you had a monopoly on local media before the internet, you don't have one after the internet.
But that doesn't mean you can't do it with a purpose.
In many cases this requires an act of altruism. To break the monopoly, somebody has to spend non-trivial resources to reproduce the ability to do something that currently only the monopolist can, without hope of turning a profit from it. People can do this if they choose to.
And the work required to do it can be distributed throughout a community.
Today we complain about the monopolistic power of, e.g. Google, who does not actually have a monopoly on search (there are many other providers), and the ability to search the entire internet isn’t really that necessary, or Apple’s mobile marketplace, where they again do not have a monopoly (people choose to develop for Apple devices, and part of choosing to sell something somewhere is agreeing to the market’s terms - as has been the case for all of time), and this is mostly to power manipulative microtransactions for kids on games. These are entirely different ballparks.
The customer is the app developer or the advertiser -- the startup Apple is extracting 30% from, or the small business which just wants to get its name out but Google systematically excludes small websites from organic search results and requires you to pay a huge chunk of your revenue to be seen.
If your prospective customers have iPhones, you need Apple's permission or you're out of business. If your prospective customers use Google search, you pay them for advertising or you're invisible. You don't get to choose which phone or search engine they use, and they rarely if ever use more than one -- unlike retail stores where it's common for the same customer to shop at dozens of different stores.
If Walmart won't carry your products in their stores, you can sell them through Target and Amazon and a zillion other retailers. The same customers can get it from another store. If Apple won't carry your app, those customers have no other viable means to get it from you.
This isn't just "manipulative microtransactions for kids on games" -- it's the reason you can't use cryptocurrency or any competing payment network for in-app purchases. Are payments not an important industry for which we care about competition?
Apple basically murdered Tumblr by giving them an impossible choice between banning adult content (causing them to lose too much of their user base) or being banned from iPhones (causing them to lose too much of their user base). Regardless of how you feel about adult content, they have the power to do this over anything. Should they be able to destroy your social network or force you to censor if they don't like your politics? How does this not matter?
If not, that would change things.
Without Stallman releasing an official Non-Commercial GPL (NCGPL) and allowing GPL'd software to be upgraded to it like it can be upgraded to AGPL, there's no way to merge non-commercial licensed code into GPL code without relicensing even if the non-commercial code is otherwise using a standard FOSS license.
>For example, if Uber’s driver rating could be exported to other similar ride-hailing services, it would have the effect of reducing the lock-in effect of Uber platform.
People are not using Uber because they don't have a rating on newcomer apps. They're using it because Uber has a large number riders and drivers in many locations, and people are aware of it.
> Mandate the use of open standards and open source competitor: All solutions must have open source alternatives before being implemented. For example, the Bavarian government switched the government operating system to Linux in 2018 to avoid being subject to Microsoft’s monopoly on software. While the project was not entirely successful
They discovered that there's a difference between "software" and "a software implementation". A myopic view of software as just code that you can throw over a fence is missing the entire reason that Windows was and is dominant on desktop. Deploying, integrating, maintaining, and operating software across an organization is labor intensive. Microsoft doesn't "compete with Linux", they compete with Red Hat or Canonical.
I think the general point is that you want to move the network effect outside the control of a single entity. So for example, like email rather than like Facebook. You relegate the service into something narrower, like payment processing, which then becomes fungible. You find the driver you want in a decentralized system and then pay them using Uber or Paypal or CashApp or Mastercard or physical cash or whatever you like.
> Deploying, integrating, maintaining, and operating software across an organization is labor intensive. Microsoft doesn't "compete with Linux", they compete with Red Hat or Canonical.
That depends on your scale, but there are different ways to go about it.
If you're small, just compiling something from github is a reasonable thing to do. You're not deploying it to a thousand workstations or trying to centrally manage anything. What do you need with Active Directory?
If you're large, you have a well-staffed IT department and possibly even your own software developers who can handle deployment issues in-house. Google has its own in-house Linux distribution in addition to ChromeOS.
The target market for Windows and Red Hat and Canonical is medium-sized entities with enough money to pay someone else to solve their problems but not enough to bring that expertise in house.
But many corporations and governments are large enough that they'd be better off doing what Google does, and open source communities can operate at the same scale. Medium sized companies that actually have a dedicated IT department could very well be better off to make a minor contribution to a community effort, and use it themselves, than lock themselves into one of these enterprise vendors. That would require more organization than is currently present, but not necessarily more than is humanly possible.
But how? Who is going to build, maintain, and provide equitable and effective governance, support, and regulatory compliance for that system? And why are others going to use it?
> If you're small, just compiling something from github is a reasonable thing to do.
Most small orgs don't have IT expertise. They buy off the shelf hardware and software. The computers they buy come with an OS on them, and they use that one. What would compel them to install a different one?
> If you're large, you have a well-staffed IT department and possibly even your own software developers who can handle deployment issues in-house. Google has its own in-house Linux distribution in addition to ChromeOS.
Google is a software company, of course they develop software. Most organizations are not software companies. There's really no reason for other large companies like PepsiCo to be making their own Linux distros for their office staff, any more than it makes sense for Google to start bottling their own beverages for their cafeteria. Yes, many are large enough that they could -- but they'll never do as good as a software company does, because they aren't one.
The question isn't whether they could do it, the question is: are there enough benefits to outweigh the costs and risks? Are you really better off training and supporting your office staff to send out invoices using SumatraPDF on Ubuntu, just to save money on Adobe/Windows licensing? For companies that have nothing to do with software, this is a recipe to lose more money in operational problems than they save in licensing.
Who does it for Debian or Blender or DNS or SMTP?
> Most small orgs don't have IT expertise. They buy off the shelf hardware and software. The computers they buy come with an OS on them, and they use that one. What would compel them to install a different one?
The computer they bought came with Windows 7, but Windows 7 isn't supported anymore. Windows 11 says it doesn't support that hardware. Meanwhile it's a quad core with enough RAM and an SSD and even if it's a few years old it's still fine. All they need it for is to use Google Docs and Chrome/Firefox. Why spend hundreds of dollars on a new machine when you can just have your niece install Debian?
> There's really no reason for other large companies like PepsiCo to be making their own Linux distros for their office staff, any more than it makes sense for Google to start bottling their own beverages for their cafeteria.
There is though.
Bottling beverages is a low margin commodity market. Software isn't. If you can bottle your beverages just as well as Pepsi, you might save a fraction of a percent on something that costs you a fraction of a percent of your revenue. If you can cut Microsoft out of a large enterprise, you can eliminate tens of millions of dollars in licensing costs. That pays for a non-zero number of software developers.
Meanwhile free software exists, so you're not starting from scratch with an empty plot of land and have to figure out how to mold plastic into beverage containers. You start with a Linux distribution that other large enterprises are already using and make some minor modifications to tailor it to your needs.
> Yes, many are large enough that they could -- but they'll never do as good as a software company does, because they aren't one.
In many cases they are. If you're a bank, you're "not a software company", but how are your customers interacting with you in the main? How about a health insurance company? A telecommunications company? An online retailer, like Amazon (which knows this quite well)?
You don't have to be "a software company" for software to be a major part of your business.
> Are you really better off training and supporting your office staff to send out invoices using SumatraPDF on Ubuntu, just to save money on Adobe/Windows licensing? For companies that have nothing to do with software, this is a recipe to lose more money in operational problems than they save in licensing.
It's not just about licensing. Okay, so you save a couple hundred bucks per seat on licensing. That's several million dollars, you could justify that, but maybe it's not much better than breakeven.
But then you're not trapped in a Skinner box, where every day you have to keep clicking "no" to a box that, if you ever click yes, you end up violating a privacy law and putting your company in regulatory peril. Or locking yourself into a vendor who can then raise prices on you by more than your original estimate of what they would cost, but by then it's too late and the cost of extricating yourself is now more than the cost of paying the danegeld.
There is value in opting yourself out of the enshittification of major platforms.
That’s how all articles or persuasion techniques work! You have a message you want to convey in mind, and then you decide how to convince the reader and write accordingly!
I am personally of the mind that we need to somehow shift the care not from personal future but to collective future. I don't have any thoughts on how to actually do that, though. It is very hard to convince people that are legitimately working hard for their personal well being that they should also give thoughts to those that come next.
What have you tried?
> It is very hard to convince people that are legitimately working hard for their personal well being that they should also give thoughts to those that come next
Haven’t read it but I keep meaning to.
I think the way to do this is to put their gains in jeopardy. We need to sit down and find an example where some money-making endeavor made life worse for most people, and scrub those dollars from the record. Perhaps we'll do this yearly.
Make dollars gained doing obvious good more reliable than dollars gained doing obvious harm.
If you make that call in a way that disagrees with your wider community, you may end up in a situation where you can't find anyone else who will accept that money, so I expect some degree of consensus will emerge.
Making that judgement call on every transaction would be tedious, so I imagine organizations would form which are trusted to make that call, and then you'd sort of "subscribe" your wallet to their decisions so that you don't have to think about it all the time.
As for regulatory capture, if such an organization stopped aligning with your interests, you could just unsubscribe from their judgements and either do it yourself or form a competing one.
And then what? Just sitting down and finding out that some thing some people did made some other people's life worse won't put the gains of the first group of people in jeopardy — unless you do something more tangible. What do propose that should be?
> If people show up at the grocery store with it
What if those grocery stores are actually the chain that's owned by one of those "bad" companies? Those grocery stores won't implement your policy, and now what? Plus I'm fairly certain you'll run afoul of some anti-discrimination law or another.
Really, if you're engaging in wishful thinking you may as well just wish for everyone to be nice and be done with it.
I disagree. "Everyone be nice" is not something that can be engineered. As hackers, we can live it, and we should, but we can't hack towards it.
> Strikebreaking is a thing.
The kind of information hygiene that makes strike breaking a more difficult endeavor is the kind of thing we can hack towards.
So many of our problems today come from how we've organized our information infrastructure (there's a lot of centralized authority in it). If we find ways to do it differently, if we practice ways to do it differently, then we may be able to achieve different outcomes.
But some of the rules that made PBS PBS have been rolled back so I don’t know that this will happen any time soon. In its place we have various foundations but those aren’t enough.
American/Worldwide capitalism incentivizes monopolist behavior
There are no currently coherent checks on this because of regulatory capture. There exist no socialist governing bodies that have power at the same scale as capitalist ones. (Note: please no “might make right” rejoinders)
The us government has set precedent over the last ~40 years which has intentionally, and structurally enshrined the capital class as legally distinct and more protected in practice
A relevant link will suffice and be greatly appreciated :-)
A pretty strange choice anyway, given that Maven Pro doesn’t even seem to be a cursive font at all.
If you are building, then it better be your core competence. If you buy, then as you become larger and more successful, buying will be more and more expensive. Its hard to make your suppliers exclusive, and competition will be somewhat maintained.
The Linux Kernel is a great example. There used to be a ton of kernels to choose from maintained and developed by small businesses, but over time Linux was chosen as a good free option. Companies would then put relatively minor investments into customizing or maintaining Linux, avoiding needing a proprietary kernel. After 20 or so years of this, there is almost no money to be made as a small operating systems vendor.
If they're using some common community-developed free software, that's good, because they contribute to it without controlling it, and then anyone else can use it too and it's one less component of the stack that someone else has to build or buy in order to compete with them.
Is there some benefit to having compiler vendors or xml libraries guarded by paywalls? This is non-obvious to me. The days before the FSF and gcc were considered "dark ages" to programmers, being locked in to a small set of proprietary of compiler vendors and toolchains, fragmentation everywhere with poor portability and poor hardware compatibility. Yet you paint this as somehow a good thing and I'm struggling to see why.
I'm also still not seeing how the openness helps establish tech monopolies, which was the main point I took issue with.
It's deeply ironic that in the end OSS, that lovely near-communist idea helped create what is probably the most successful capitalistic business models of all time... but, the fact that these monopolies are still there and unregulated and unharmed, that's probably a US-specific political failure and cannot be fixed with technological hacks and definitely not by one of their core enablers (OSS).
Well said. Hopefully the US Government will start being harder on monopolies in some meaningful way.
Wikipedia says the goals of the GPL are "the four freedoms to run, study, share, and modify the software."
It's a pyrrhic victory in today's world where the building blocks are open source, but the full stack is not, and the software is running on corporate servers where users have no control over it.
Companies have simply commoditized and devalued the software and started hoarding the far more valuable user data instead.
If they need more capacity than 1M would plausibly buy them from themselves, just establish however many more identical companies are necessary, none of which exceed 1M$.
In practice though amazon engineers would reverse engineer any tool they cant legally use and need.
Meanwhile, you propose to pay nothing at all?
Also, may I ask what is your fundamental problem with such a license besides perhaps the name and purity? This license will only affect the insanely rich, and in a way that they will most likely not even feel.
That's the practical concern. The idealistic concern is that free software is all about handing power to the users. Much like how a pen can't restrict how you use it, developers shouldn't be able to restrict how users can use it. Note that the only restriction GPL has is that you cannot further restrict it (tolerance paradox). It's an ethical position to give power to the users, regardless of who they are, because software owners should not have the power to control the users. From here, a software developer decreeing "you may not use my stuff if you make more than X amount per year" is abusing their power over users, which is unethical. Imagine if a pen company told you you cannot use their pens if you met X Y or Z conditions. It's just not right.
Stallman covered this: https://www.gnu.org/philosophy/programs-must-not-limit-freed...
Free Software is supposed to create a gift economy. People create and release their work in good faith, others copy and improve, and contribute back to the community. Gift economies work if most continue to gift their work to society in good faith.
If some actors decide to largely take and but rarely give out of self-interest, or worse use their gifts to cement their economic position, then the gift economy turns into an exploitation economy. The way to fix such an economy is to intentionally exclude those who are exploitative.
Free Software failed to create a gift economy. I am all for abandoning it in favor of the Guarded Free Software economy as you propose.
In a gift economy. I gift you something. You gift someone something. That someone could be me, but if I don't need/want anything you need, then you just gift to someone else. If there are gifts going around all the time, all in good faith, then most people will receive what they need.
You see gift economies in families or friends all the time. People help others according to their abilities. My rich friend takes me out to an expensive dinner, poor me gets us lunch from that famous food truck. I nurse my uncle when he breaks his leg. He lets my brother's son to live with him for free during college. There is no accounting of how much anyone gave. Just give what you can.
I recommend reading Cory Doctorow's novel Walkaway which describes gift economies beautifully.
It doesn't solve the problem of someone being able to operate the software on a server, where the users have no control of it and their data is held hostage, too.
The issue today is that people want more than open source ever promised. The source is available, you can make your own changes, you can distribute them or even start a business but that's not enough for most people/companies. They want a ecosystem, a non-profit organizing the effort, ability to buy support, etc. None of that were in the initial goals.
Either they are BSD/GPL and accept competition or they do not, and they are not open source.
The only update that I agree with is the AGPL because of the nature that software is being closed back up and not shared for studying purposes.
Only if you’re not in the SaaS space. If you are, then open-sourcing your tech is a risky strategy because then if you have any success, you start to compete with bigger players who take what you’ve invested & then it’s a game of $ which is harder to compete with.
If you are a software as a product company and you open source your software, someone else may put it into a box and sell it beside you.
If you are a software as a business process company and you open source your software, some else may build into their business and compete against you in whatever your marketable output is.
If you are a software as a service company, and you open source your software, someone may build a service that is remarkably similar to yours.
Open source is great for things your company depends on --- maybe you get community improvements, and you can build and change it how you like; but if the core of your business is open source, that's a choice that makes it very easy for competition to form. OTOH, if you want to build something that other companies depend on, and there's open source alternatives, it's going to be hard going; I wouldn't use a closed source web server or database server or search engine, given there are good open source options; although maybe if those options don't work out, I'd consider a closed source vendor.
You really can't in the original sense, though. A person using Facebook, Spotify, or YouTube can't make any modifications to those products whatsoever despite a lot of the code being "open source". It was never about taking somebody else's code and making a business.
More disappointing, users often don't even have direct access to their own data on those services.
Correct.
At this stage the weaponization/enshittification of FOSS by abstracting its original goals, and making it about the equivalent of leverage for larger firms to use for cheap labor/R&D will only likely change if people only use reciprocal licenses with clauses for SaaS firms to straighten up (e.g. AGPL v3) AND legally enforce them in the courts.
Ideally peer-based but I know federation is much more popular and that can work also.
But it's not just the fact that they leverage open software to build their closed systems, it's the large networks and marketplaces they have. The only way to compete with that is through decentralized systems. Which (despite the connotation of the name), are the best approaches for creating large scale, open, holistic networks.
Moving to decentralized protocols also means it is harder to dominate with one piece of software. Ideally the protocols are in some kind of protocol registry so they can evolve freely.
But on some level I believe that some entity like a government needs to mandate this. Although I know there are many people who are philosophically against governments or practically against them due to their track record of being generally awful. But at some level you may have to force people to cooperate in some basic technical way in order to avoid the technopolies.
E.g. TCP/IP is open; that doesn't make the Internet and Web open.
Open formats and protocols can serve as monopoly-enabling catalysts.
Could copyleft protocols be a thing?
We don't need copyrighted protocols and APIs.
Make an app for a decentralized protocol, add some polish and maybe give something away with it which is cheap at scale (storage, bandwidth) as the loss leader and establish dominance.
We just need to mandate the basic / baseline technology. External File System that is supported everywhere ( ExFAT ? ), PDF, Word, Excel / Spreadsheet, Archive Files, Photo/ Graphics, Audio and Video.
Unfortunately I dont it will ever happen. One being goverement doesn't think like that. Two even if they do they dont know which one to pick. And three since they cant pick anyone they will decide to make their own.......
Right now I really wish we could create something on top of the soon patent free AAC-LC and H.264 Video Codec.